Digital image input devices such as digital still cameras and scanners have become very popular in recent years, and image output devices typified by ink-jet printers also have been improved in resolution and lowered in cost. This has led to an environment in which a user can print images at home. Meanwhile, many techniques relating to image correction have been proposed and improved.
Many image correction methods rely upon a fundamental concept that involves what method to use to correct highlight and shadow points of an image. That is, there is an approach that corrects an image upon judging the direction of a color balance correction from a highlight point or shadow point. Furthermore, since a color balance correction from a warm color system of a specific scene typified by the sky at sunset to a cool color system falls within an undesirable correction, the scene is evaluated based upon various parameters and the correction is not applied to an image of such a scene.
The color reproduction performance of image input/output devices has been improved greatly and now even very slight changes in color ascribable to a color balance correction are recognizable. As a consequence, there are cases where a specific color shift, such as a color shift in the yellow or blue direction, gives a disagreeable sensation when applied with regard to a slight color change in the color of the human skin, which is a memory color. Further, from the viewpoint of desirability, it is important that a color balance correction especially from a warm to a cool color system be suppressed in an image that has undergone exception processing based upon the above-described scene evaluation.